System and method for longitudinal assessment of a biomarker, and mri apparatus

Abstract

A system and a method for evaluating a longitudinal evolution of a quantitative biomarker measured for a biological object include receiving longitudinal measurements of the quantitative biomarker for the biological object. A statistical parameter, which characterizes a variability of the acquisition technique being used, is calculated for each measurement. The calculated statistical parameter is used for each obtained biomarker value for generating synthetic data having a distribution which follows the statistical distribution of possible values for the biomarker. The sampled values are fitted by using a fitting function for each bootstrapping fitting. A fitting parameter of the fitting function is extracted for each fitting. The longitudinal evolution of the quantitative biomarker is evaluated by statistically comparing the extracted fitting parameters against a reference value.

1 . A method for evaluating a longitudinal evolution of a quantitative biomarker measured for a biological object, the method comprising:
 receiving, for the biological object, longitudinal measurements of the quantitative biomarker, N measurements M_1, . . . ,M_N, being received, each measurement M_i, with i=1, . . . ,N, being performed at a timepoint T_i according to an acquisition technique A_i, when i≠j, then T_i≠T_j∀i,j∈{1, . . . ,N}, for each measurement M_i, and a value V_i being obtained for the biomarker;   calculating, for each measurement M_i, at least one statistical parameter characterizing a variability of the acquisition technique A_i being used, the statistical parameter being a measure of a statistical distribution of possible values for the biomarker when carrying out the measurement M_i according to the acquisition technique A_i;   for each obtained biomarker value V_i, using the calculated statistical parameter for generating synthetic data having a distribution following the statistical distribution of possible values for the biomarker, each synthetic data representing a simulated value for the biomarker and being assigned the timepoint T_i at which the measurement M_i was performed;   performing a plurality of bootstrapping rounds, for each bootstrapping, one or a plurality of values among the measured value V_i and the biomarker simulated values generated for the measured value V_i being randomly sampled for each timepoint T_i;   for each bootstrapping, fitting the sampled values by using a fitting function;   for each fitting, extracting a fitting parameter of the fitting function; and   evaluating the longitudinal evolution of the quantitative biomarker by statistically comparing the extracted fitting parameters against a reference value.   
     
     
         2 . The method according to  claim 1 , which further comprises selecting the biomarker as:
 a physical characteristic of the biological object or of a region of interest within the biological object; or   a measurable physiological characteristic of the region of interest or of the biological object; or   a measurable biochemical characteristic of the region of interest or of the biological object.   
     
     
         3 . The method according to  claim 1 , which further comprises selecting a lesion volume as the biomarker, the acquisition technique A_i including segmenting lesions in the biological object by using a segmentation algorithm, and the statistical parameter being a measure of a variability of results outputted by the segmentation algorithm. 
     
     
         4 . The method according to  claim 1 , which further comprises calculating the statistical parameter by performing test-retest experiments for the acquisition technique A_i. 
     
     
         5 . The method according to  claim 1 , which further comprises carrying out the statistically comparing of the extracted fitting parameters against a reference value by calculating a median estimate of the fitting parameters obtained for the biological object and testing the median estimate against a null hypothesis. 
     
     
         6 . The method according to  claim 1 , which further comprises configuring the fitting function for fitting a first set of consecutive timepoints T_i independently from a second set of consecutive timepoints T_i, distinct from the first set, to independently evaluate, for the first set and the second set, a respective longitudinal sub-evolution of the quantitative biomarker. 
     
     
         7 . The method according to  claim 1 , which further comprises selecting the acquisition technique A_i as an MRI acquisition technique or an X-ray acquisition technique, or a computed tomography acquisition technique or an ultrasound-based acquisition technique. 
     
     
         8 . The method according to  claim 1 , which further comprises selecting the fitting function as a linear regression. 
     
     
         9 . The method according to  claim 1 , which further comprises:
 (a) ∀i,j∈{1, . . . ,N}, providing the acquisition technique A_i to be the same as the acquisition technique A_j, or   (b) providing at least one acquisition technique A_j being different than the acquisition technique A_i, with i≠j and i,j∈{1, . . . ,N}.   
     
     
         10 . The method according to  claim 1 , which further comprises providing the reference value as either a value, or a distribution extracted from a set of measurements M′_1, . . . , M′_K, called a reference set, each measurement M′_r with r=1, . . . , K, with K>1, having been performed at a timepoint T′_r, and the measurements M′_1, . . . ,M′_K having been performed for the biological object, or for another biological object. 
     
     
         11 . A system for evaluating a longitudinal evolution of a quantitative biomarker measured for a biological object, the system comprising:
 a first interface for receiving or acquiring longitudinal measurements of the quantitative biomarker for the biological object, N measurements M_1, . . . , M_N, being received or acquired, each measurement M_i being performed at a timepoint T_i according to an acquisition technique A_i, when i≠j, then T_i≠T_j∀i,j∈{1, . . . ,N}, and for each measurement M_i, a value V_i being obtained for the biomarker;   a memory for storing the N measurements M_i and their respective associated biomarker value V_i; and   a control unit including a processor, the control unit being configured for carrying out the method according to  claim 1 .   
     
     
         12 . A Magnetic Resonance Imaging apparatus, comprising the system according to  claim 11  configured for performing the N measurements.